Nitrogen ion implantation is used industrially to increase the surface hardness and wear resistance of metals, which can result in a tremendous increase in the lifetime of tools. This process does not require the elevated temperatures used for thermal diffusion of nitrogen into metals. In addition, it is not a coating, so it has no adhesion problems. The implantation process also has the effect of producing much smoother surfaces than for untreated material, resulting in less friction for contacting surfaces such as ball bearings. Deep implants are preferred, which makes the implantation of atomic nitrogen ions (N.sup.+) rather than molecular nitrogen ions (N.sub.2.sup.+) necessary, since for a given acceleration energy N.sup.+ ions are implanted deeper.
The best wear and corrosion resistance is achieved when implantation is done with only N.sup.+ ions, rather than with a beam having both N.sup.+ and N.sub.2.sup.+ ions, because the N.sub.2.sup.+ ions will have half the desired energy and will be shallowly implanted, resulting in a poor control over the implanted depth.
Ion sources providing a high current density are highly desirable since less time will be spent per part treated for the same ion dose. Nitrogen ion implantation is usually carried out at energies of 10-400 keV and dose levels of 10.sup.16 -10.sup.18 ions/cm.sup.2.
Prior to the present invention, the ion sources available for ion implantation of materials had ion beams with too great a percentage of N.sub.2.sup.+ ions as to enable the ion beams to be used directly. In order to remove the undesired N.sub.2.sup.+ ions and produce a N.sup.+ beam of sufficient purity for industrial application, mass separation procedures are used. These processes use a large magnet near the extracted beam to bend the paths of the ions in the beam. The paths of the N.sup.+ ions will be bent more than the paths of the N.sub.2.sup.+ ions, thus enabling a beam to be produced having only N.sup.+ ions.
A mass separation process imposes significant limits on implanter design by requiring relatively low energy in extraction so that a magnetic field of reasonable strength can provide sufficient bending for separation. Further, the need for the magnets downstream of the ion source will increase the length of the beam path to the material to be treated with consequent losses in energy. Further, since the beam has an appreciable cross-sectional area as it passes the magnet, the paths of the N.sup.+ ions closes magnet will be bent to a greater extent than the paths of the N.sup.+ ions further from the magnet such that there is an undesirable diffusion of the resulting N.sup.+ ion beam. Elimination of the mass separation step would allow a simpler, more efficient, more compact implanter, with greater throughput of implanted parts.
A multicusp ion source capable of producing beams with greater than 90% N.sup.+ ions has been described in S. A. Walther, K. N. Leung, and W. B. Kunkel "Production of atomic or molecular nitrogen ion beams using a multicusp and a microwave ion source," J. Appl. Phys., 63(12), pp. 5678-5682, Jun. 15, 1988. Even though the percentage of atomic N.sup.+ ions in the extracted beam is relatively high, the purity is still not sufficiently high as to enable ion implantation processes to be carried out without a mass separation process to remove the molecular N.sub.2.sup.+ ions from the beam. Further, the ion beam current density available from the described device is too low and the operating pressure is too high for practicable commercial operations.